GM’s Cruise robotaxi division showed off their work on adapting their custom robotaxi, known as the “Origin” for wheelchair use. The vehicle has a built in ramp, can kneel and has wheelchair docking options replacing a classic seat.

While the potential of robocars for the disabled has often been mentioned, some disabled advocates have been disappointed that nothing is yet deployed. They will be pleased to see this. Waymo’s very first member of the public to ride was Steve Mahan, director of the Santa Clara blind center — I participated in making a video about this. For now, though, Waymo says that disabled riders will be redirected to a human driven car. The potential for much more exists. This includes

  • Dedicated vehicles for wheelchair users
  • Robotic Paratransit — which may offer an early revenue opportunity for robotaxis

Dedicated vehicles

General practice in access it to demand that all vehicles be accessible, so that disabled passengers do not risk becoming 2nd class passengers. This makes sense when you talk about things like a bus — nobody wants to find that the bus is not accessible and they must then wait for a later bus which is, or to get directed to a completely alternate mode if it involves more wait.

With dispatched vehicles, like Robotaxis, it is possible to adapt only some of the fleet, because you can have a mixed fleet and send the right vehicle for each trip. Future robotaxi fleets are likely to consist of both small vehicles for 1-2 passengers, and larger ones. If a request comes for a solo trip—and that’s about 80% of all trips—you can send a small vehicle which is lower cost and uses less road space and less energy. That’s a win for everybody and these vehicles should eventually make up most of the fleet. When 4 people need a ride together, something more car sized makes sense, and larger groups can use van-sized vehicles.

Cruise’s special Origin replaces a standard seat with a wheelchair dock though it may also be possible to have a seat transform so the space can handle both.

While having access in all vehicles is often the preferred plan, a mixed fleet can actually produce superior results for the disabled. Access modifications can cost quite a bit of money, and as such there is only so much budget for them. With a fleet of 1,000 vehicles, the budget per vehicle is more limited, while in a sub-fleet much more can be spent per vehicle, both to improve the quality of the service for the disabled, and to add more vehicles to the fleet to improve service levels. For example, if basic improvements cost $1,000 each, then a million dollars is needed over the whole fleet. However, if only 100 vehicles are adapted, there is $10K available for each one. Alternately, it can be possible to adapt 100 vehicles and simply add another 50 adapted vehicles , so that the accessible section of the fleet is overprovisioned for the needs of the disabled, and their average wait times for the special vehicles are shorter then general average wait times—separate but superior, rather than separate but equal or inferior.

One type of vehicle that might be of interest is the small, hallow vehicle. Such vehicles are already popular in the Netherlands for manual driving, and include things like the Canta, and the American Kenguru (prototype) and others.

Hollow cars such as these are of course lower cost, use less energy and can also be much faster to enter and exit. These designs do not support travel with a companion but other designs can do so. They are not currently adapted for self-driving, but this is a reasonable future plan for them.

The right fleet will contain vehicles that can carry solo disabled passengers, such passengers with companions, solo walking passengers and groups of all types of passengers. Ideally the right size vehicle is sent, but a larger vehicle can always be sent if that is what is available or closest.

Overall these efforts should result in a tremendous increase in mobility for people with all sorts of disabilities. The main group they won’t help are those who need assistance with luggage or cargo and don’t have a companion—those passengers may continue to request vehicles with a human driver or human assistant.


A special opportunity exists in the world of paratransit, which is complex and heavily regulated and not very satisfactory to any of the parties involved. The ADA requires all transit agencies to offer a parallel transit offering for the disabled to those who live within 3/4 mile of a transit line but who can’t readily use the transit line, either because the transit vehicles or stations are not accessible, or the passenger’s disability makes it challenging for them to travel to the stop.

The paratransit offering tends to be door to door, and may cost no more than double a regular transit fare. However, the agency can require that rides be booked a day in advance, with a one-hour pickup window and shared travel. They often do this because as you might expect, door to door ride service (like a taxi) costs a lot more than the $5 they can charge, and so they need to discourage overuse to keep the budget under control. The result though, is that the average paratransit ride in the USA rose from $30 to closer to $50 in the previous decade. In New York City, it’s over $100. As a result, paratransit uses around 10% of the $80 Billion spent each year on public transportation in the USA. To serve that market efficiently would handily pay back the billions companies like Cruise have invested in robotaxis even if they never carried an able-bodied passenger.

Some transit agencies, tired of this disparity, have even allowed passengers who can use a taxi do so, because they lose a lot less money on that trip than sending one of the expensive paratransit vans. Some paratransit passengers (particularly those with mental issues like dementia) travel with a companion, and the rules allow such passengers to bring companions. For some passengers there may need to be additional service, but a large number of current disabled passengers could travel in a vehicle such as the Origin above. In addition, they could in most cases get same-day, door to door service, with more reasonable pick-up windows, while still receiving service at a much more affordable cost to the city. Some passengers would need a human attendant, with higher costs, but that attendant need not drive, and those passengers could receive the more standard group/day-ahead service level they get today (or more likely much better) to control costs.

Costs could also be reduced with robotic first-mile services to take suitable passengers from even more than 3/4 miles away (as many agencies already support) to accessible bus/train stops to ride with other members of the public. Indeed a suitable vehicle with lift might be able to take a passenger onto a bus that’s not otherwise easy to climb into.

Paratransit moves only a small fraction of those with travel-limiting disabilities. Passengers must get certified as having a qualifying disability to use it. Regular robotaxi service should work very well for people with modest mobility issues (such as need for canes and crutches) and sensory problems (like vision impairment.) As robotaxi costs start dropping to approach the cost of private car travel, they will make transport available to a major segment of the population who can’t drive (or afford) a private car or find it painful to take transit, but who don’t qualify for paratransit or can’t stomach its need to be book in advance.

Sadly, the San Francisco Municipal Transit Agency has decided to pick a fight with robotaxi companies like Cruise and Waymo. San Francisco spends about $30M per year to provide 800,000 rides to 14,000 disabled residents, about $37 per ride. I suspect Cruise and Waymo could cost the cost of that at least in half, and perhaps to 1/4, but it remains to be seen how friendly they will be with the agency when the time comes. Too bad, because the rest of the agency could certainly make use of those savings, or for that matter, so could the taxpayers or the disabled.